Video Motion Vector Correction for Boundary Block Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video signal processing methods lack efficiency in coding, particularly in handling motion information and boundary prediction areas, leading to suboptimal compression and decoding performance.
Innovation Solution
A video signal processing method that corrects motion information using techniques such as motion vector difference, merge mode with MVD, template matching, bilateral matching, optical flow-based TM, and decoder-side motion vector refinement, and applies adaptive motion vector resolution and deblocking filters to improve coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion information is acquired from neighboring blocks without correction, then decoding complexity is reduced, but coding precision deteriorates
Solution Approach 1:
The patent implements feedback mechanisms where decoded motion information from neighboring blocks is corrected using refinement techniques (DMVR, AMVR) and fed back into the decoding process. This iterative correction improves motion information accuracy while managing complexity through selective application of refinement methods based on block characteristics and mode types.
Solution Approach 2:
The patent applies preliminary correction actions by preparing refined motion information from neighboring blocks before using it for current block decoding. Techniques such as template matching and bilateral matching are performed in advance to correct motion vectors, ensuring accurate motion compensation without increasing real-time decoding complexity.
2Measurement precision
If adaptive motion vector resolution is applied to all blocks, then motion information accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent applies adaptive motion vector resolution selectively based on local block characteristics such as motion complexity, block size, and prediction mode. High-resolution motion vectors are applied only to blocks requiring precise motion compensation, while simpler blocks use standard resolution, optimizing the balance between accuracy and processing complexity.
Solution Approach 2:
The patent implements dynamic adaptation of motion vector resolution where the resolution level is adjusted based on decoded block characteristics and motion activity. The system dynamically selects between different resolution levels (e.g., 1/4 pixel, 1/8 pixel, 1/16 pixel) to match the actual motion complexity of each block, avoiding unnecessary processing overhead.
3Measurement precision
If boundary blocks are processed without padding, then processing speed is maintained, but prediction accuracy deteriorates
Solution Approach 1:
The patent performs preliminary padding of boundary blocks by generating virtual reference samples from available picture data before boundary block processing. This pre-computation ensures that boundary blocks have sufficient reference information for accurate prediction without requiring complex runtime handling, maintaining processing speed while improving accuracy.
Solution Approach 2:
The patent introduces padding reference blocks as intermediary elements that bridge the gap between actual picture boundaries and prediction requirements. These padding blocks contain synthesized reference samples that enable accurate motion compensation and filtering at picture boundaries without accessing non-existent external data.
4Measurement precision
If deblocking filter is applied to all boundaries, then prediction quality is improved, but processing overhead increases
Solution Approach 1:
The patent applies deblocking filtering selectively based on local boundary characteristics such as block size, prediction mode, and motion activity. Stronger filtering is applied to boundaries with visible artifacts or high motion complexity, while boundaries with smooth transitions use minimal or no filtering, optimizing quality while reducing processing overhead.
Data Source
AI summary
A video signal decoding device is disclosed. The video signal decoding device comprises a processor, and the processor obtains motion information from neighboring blocks of a current block and corrects the motion information to obtain motion information of the current block.